NEED HELP!

[NyiriOwner]

Okay that's good because then yours has a TDP of 375 Watt and should definitely be able to run of of your 750 W PSU

 

That's probably the problem. If im not mistaken they start throttling at 100c. And anyway 90c is not a temperature you want them too run at for extended periods of time. Try running them without a side panel on your case. If they preform better then that's the problem.  Also you might wanna check with a program like Msi afterburner (or the sapphire equivalent) if its lowering the clock rate when in game.

I just opened the 7990, replaced cooling paste and it went down in the 80 (+ - 80 depends on load), and none of the clockspeeds changes

[NyiriOwner]

Try using mantle in Bf4 if you haven't already. That should help to alleviate CPU bottlenecks. Mantle was terribly broken at the start, but now with the 14.4 driver you will experience much better performance. Try running the test range. Multiplayer is known for performance issues. That CPU overclock should help.

I need to download BF4 first, so i will update tomorrow about BF4 and mantle

[iExpurge]

So both the clock speeds are running at their normal rates? then i cant think of anything else that might cause it.

you have updated drivers, its not under clocking due to temperature, your PSU is powerful enough...

I think its time to contact sapphire support, and maybe get an RMA.

[PCfees]

I agree with the majority of the people thinking it may be a power supply issue. I'm currently running a 5970 Dual-GPU card, and the power requirements are more or less the same. Working with bigger, more power-hungry cards can be a drawback sometimes. Even with at 850w, the power supply is at pretty much full load when gaming for me. And since your card has similar power requirements, I REALLY don't think 750w is going to let your card really push itself to it's potential.

[NyiriOwner]

So both the clock speeds are running at their normal rates? then i cant think of anything else that might cause it.

you have updated drivers, its not under clocking due to temperature, your PSU is powerful enough...

I think its time to contact sapphire support, and maybe get an RMA.

I already had a RMA, but they just said everything was fine and returned it to me... Getting worryed i wont be able to fix this problem in a while...

[NyiriOwner]

I agree with the majority of the people thinking it may be a power supply issue. I'm currently running a 5970 Dual-GPU card, and the power requirements are more or less the same. Working with bigger, more power-hungry cards can be a drawback sometimes. Even with at 850w, the power supply is at pretty much full load when gaming for me. And since your card has similar power requirements, I REALLY don't think 750w is going to let your card really push itself to it's potential.

I will find out, will be getting a new PSU after this summer i think. Hopefully that will fix it

[iExpurge]

I will find out, will be getting a new PSU after this summer i think. Hopefully that will fix it

I already had a RMA, but they just said everything was fine and returned it to me... Getting worryed i wont be able to fix this problem in a while...

Ow that sucks then, can you ask what their test system looked like? So you compare it to yours.

 

But on the PSU side, are you sure your PSU is running properly? because a 750W should work. The rest of your system should be like this(Worst case scenario)

8350        130W

HD7990   375W

MOBO      75W

So that still leaves 180W for random stuff like HDD's and RAM which is plenty. So unless your PSU isn't working properly thats not the problem.

[NyiriOwner]

Ow that sucks then, can you ask what their test system looked like? So you compare it to yours.

 

But on the PSU side, are you sure your PSU is running properly? because a 750W should work. The rest of your system should be like this(Worst case scenario)

8350        130W

HD7990   375W

MOBO      75W

So that still leaves 180W for random stuff like HDD's and RAM which is plenty. So unless your PSU isn't working properly thats not the problem.

Some other guy said somthing about amps, and not the watts. I have no idea, but that may be diffrent or what? The 8350 is overclocked now, so i guess it takes more power, but i dont know this stuff :c Any way i can check how much power my computer uses? And any way i can check my PSU works properly? Ty for taking time to help me BTW, really appreciate it

[iExpurge]

The thing about the amps is that the most power intensive components are hooked up to the 12v rail. And as such a good psu can deliver all of its power over the 12v rail. And wattage is Amps*volts=watt so if you know that its 12v then the amount of amps is important to know because it gives you the wattage. But a calculation in an earlier post stated that this psu can deliver 62 amps over the 12v rail so that's 12*62=744 which is close enough to 750W.

You can check the amount of power draw from the wall socket with something like a kill-a-watt. I don't suppose you have another similar PSU to replace it with to test?

 

Have you checked if the clock speeds are normal? That would indicate if there is anything wrong with its power delivery or general under clocking.

[NyiriOwner]

The thing about the amps is that the most power intensive components are hooked up to the 12v rail. And as such a good psu can deliver all of its power over the 12v rail. And wattage is Amps*volts=watt so if you know that its 12v then the amount of amps is important to know because it gives you the wattage. But a calculation in an earlier post stated that this psu can deliver 62 amps over the 12v rail so that's 12*62=744 which is close enough to 750W.

You can check the amount of power draw from the wall socket with something like a kill-a-watt. I don't suppose you have another similar PSU to replace it with to test?

 

Have you checked if the clock speeds are normal? That would indicate if there is anything wrong with its power delivery or general under clocking.

This is a bronze 80 plus certification PSU, dosent that mean it cant use all its power, or am i jst retarded right now? And no, i dont have any other PSU. This is a pic after a Valley run, you can check if its normal or not.

[iExpurge]

It doesn't show the clocks on the second GPU. But the clocks on GPU 1 seem perfectly normal... maybe GPU 2 isnt working properly but that would be strange considering CF doesn't work with different clocked cards.

But im starting to think this is not something i can fix sadly.

[7850OC]

new test (yey, i am back)

 

run a GPU-Z (or two at the same time)

 

and use it to check (in real time) the real temps of the "two" gpus with NO load (and show us a picture)

 

(1000 post, yey)  B) 

[NyiriOwner]

new test (yey, i am back)

 

run a GPU-Z (or two at the same time)

 

and use it to check (in real time) the real temps of the "two" gpus with NO load (and show us a picture)

 

(1000 post, yey)  B) 

Hi again! The right one dosent show upp, cuz of the tecnology that turns off the GPU when idle, but i took with both. Tell if u wanna see anything else

[iExpurge]

Could you do that double Gpu-z again but with a form of load in windowed mode?

Because 59 degrees in idle with only one gpu running is an awfully high number.

 

EDIT: on a side note how long after the valley run did you take that picture in afterburner? Because the graph only updates every couple of moments and the current temp might have dropped while you were taking that screenshot, giving us the impression that the temps are better than they are.

[AHaskin14]

update EVERYTHING on your motherboard and gpu and cpu support drivers and see.

[7850OC]

Hi again! The right one dosent show upp, cuz of the tecnology that turns off the GPU when idle, but i took with both. Tell if u wanna see anything else

great work, now like @iExpurge say, redo the same test but add a valley run on the back at the same time

 

(use the window mode on custom preset for that, no need to run the Exteme HD preset this time)

 

 

Could you do that double Gpu-z again but with a form of load in windowed mode?

Because 59 degrees in idle with only one gpu running is an awfully high number.

 

EDIT: on a side note how long after the valley run did you take that picture in afterburner? Because the graph only updates every couple of moments and the current temp might have dropped while you were taking that screenshot, giving us the impression that the temps are better than they are.

yes, something is wrong on that no load temp  :blush:

[NyiriOwner]

great work, now like @iExpurge say, redo the same test but add a valley run on the back at the same time

 

(use the window mode on custom preset for that, no need to run the Exteme HD preset this time)

 

 

yes, something is wrong on that no load temp  :blush:

The second GPU dosent seem to take any of the load, and are at really low sppeds... Dafuq

[NyiriOwner]

update EVERYTHING on your motherboard and gpu and cpu support drivers and see.

Everything i know of is updated, i think... Im no pro at this stuff, so any way i can check?

[7850OC]

The second GPU dosent seem to take any of the load, and are at really low sppeds... Dafuq

wow  that second gpu doesn't work at all on load

 

:blush:

[7850OC]

temps on the 1s gpu are fine (80º is ok for load in there) 

[NyiriOwner]

Could you do that double Gpu-z again but with a form of load in windowed mode?

Because 59 degrees in idle with only one gpu running is an awfully high number.

 

EDIT: on a side note how long after the valley run did you take that picture in afterburner? Because the graph only updates every couple of moments and the current temp might have dropped while you were taking that screenshot, giving us the impression that the temps are better than they are.

I got the temps down to 45 (idle) while running fans at 100% for a while. Went fast up to 50, and was at 55+ in a few min. So the temps are really that high...

[iExpurge]

yes, we know the problem now. That means we can fix it . Temps might become an issue when both gpu are running at full speed, but we will worry about that later. Have you checked CCC for options regarding crossfire? There might be some power saving crap enabled. Or something else like crossfire not being enabled.

 

EDIT: missed some stuff
 

The second GPU dosent seem to take any of the load, and are at really low sppeds... Dafuq

[NyiriOwner]

yes, we know the problem now. That means we can fix it . Temps might become an issue when both gpu are running at full speed, but we will worry about that later. Have you checked CCC for options regarding crossfire? There might be some power saving crap enabled. Or something else like crossfire not being enabled.

 

EDIT: missed some stuff

 

Shouldent be any powersavings on

[iExpurge]

When in doubt check ALL the stuff and since we have no other solution yet you might as well try it.

[7850OC]

Shouldent be any powersavings on

is that card have some time of swich ? (for uber mode or dual bios)

 

you may want click that to (if is there)